Effect of electrospark deposited surface layers on bond strength of titanium–porcelain

Materials and methods

Substrate materials used were titanium alloy Ti–6Al–4V, with the following nominal composition (wt): Ti–6·03Al–4·02V–0·3Fe–0·1C–0·01H–0·05N–0·1O. Specimens were cut into 25×3×0·5 mm. The surfaces were polished with abrasive SiC papers of 240–1000 grit and ultrasonically washed in by ethanol and pure water.

Electrospark surface deposited modified layer was prepared by ESD with SQ-2 type strengthening equipment. For the treatment, the pulse power supply was employed. Nb with Φ5 mm diameterand YG8 (92WC+8Co) hard alloy and Si with Φ3 mm diameter were used as deposition electrodes. The samples as the cathode were modified in silicone oil medium or in air to prepare interlayer. Before porcelain, the materials were prepared by machinery and Al₂O₃ sand blasting of 50 μm (at a pressure of 0·2 MPa, from a distance of 1 cm, with a 45° angle). The preparation of the materials includes mechanical polishing and cleaning in acetone, ethanol and distilled water by ways of ultrasonic for several times and dried at room temperature for microstructure characterisation and porcelain application.

For casting titanium alloy specimens, thin bonding porcelain (super porcelain Ti-22 powder, Noritake) was applied to the central portion (8×3 mm) of the specimen surface. According to the protocol ISO 9693, after the porcelain was fired, 0·2 mm opaque porcelain was applied in a masking layer, as it was a common practice with the porcelain fused to metal technique, and fired. 0·8 mm dentin porcelain was applied onto opaque porcelain using a custom-made jig that controlled the position and the thickness of the porcelain.

After firing, the bond strength test was performed with a three-point bending device on a universal testing machine. The specimens were placed with the porcelain facing down in the bending apparatus with rounded supporting rods 20 mm apart and loaded at the centre with a rounded bending piston (radius, 1·0 mm). Force was applied at a constant rate of 0·5 mm min⁻¹. The fracture force F_fail (N) was measured for failure of specimen by a deboning crack occurring at one end of the porcelain layer. The debonding/crack initiation strength τ_b was calculated by the equation τ_b = k×F_fail, where the coefficient k was a function of the thickness of the metal substrate and the value of Young's modulus of the used metallic material (k = 4·7).

Low fusing porcelain for titanium was used, and the sintering conditions are shown in Table 1. For increasing sintering temperatures, the porcelain layer thickness decreases. The porcelain at 800°C generated the thinnest film. Moreover, regardless of the sintering temperatures, surfaces were least susceptible to oxidation as it was specimens.

OPEN IN VIEWER

Table 1

Sintering conditions of titanium–porcelain

Name	Starting temperature/°C	Sintering temperature/°C	Sintering time/min	Porcelain layer thickness/mm
Adhesive	500	800	3	0·2
Opaque	500	780	3	0·2
Dentine	500	760	3	0·6

Characterisation

The surface morphology and the cross-section structure of ESD modified structures were evaluated by SEM (JSM6460, JEOL, Japan). The element composition analysis of the fractured face was evaluated by an energy dispersive X-ray spectroscope (INCA-300, Oxford, UK). X-ray diffraction measurements were carried out on a D/max-2200pc X-ray diffractometer (Rigaku, Japan) Cu K _α (λ = 0·154 nm) radiation at 40 mA. The diffractograms were recorded in the 2θ range of 20–90° with a 2θ step size of 0·02°. The microstructure of the coating layer was analysed by the glow discharge spectrum on a GDA-750 glow discharge spectrometer (Germany). According to ASTM G54-77, a high temperature oxidation experiment of the ESD layer is also carried out, which is a static oxidation test. The sample is placed in the resistance furnace and oxidised for 100 h. The loss of weight was measured in a high precision electronic analytical balance (Shimadzu AUW220D).

Performance of modified layer by ESD

Figure 1 represents a cross-section of a control specimen before and after sandblasting. It indicated that the electrode material transferred to the titanium alloy surface and formed complex compounds in the discharge process. Those compounds, derived from electrode and substrate materials, transformed the strengthening point under the cooling effect and then the electrode surface coated at the substrate surface to form a strengthening layer with the electrode repeatedly discharging gradually. The results show that the ‘melting’ characteristics of deposited layers prepared by Nb/YG8 electrodes in the air and Si electrode in the silicone oil are obvious, and their roughness of deposited layer is small. However, deposited layers prepared by Si electrode in the air is increased (R_a is 1·45 μm) and show globular fungi form characters. It is not only attributed to the splash degree for the reactants in the air and the silicone oil medium, it is but also relative to the thermal pulverisation and mechanical wear for Si electrode in discharge treatment process. In addition, the modified layer prepared in the air is easier to crack than that in the silicone oil, which is a common phenomenon in the deposition of titanium electrospark.

OPEN IN VIEWER

1

a before sandblasting; b after sandblasting

General properties of different ESD layers on the titanium alloy surface are shown in Table 2. Compound compositions of intermediate layer were confirmed by X-ray diffraction results analysis. The thicknesses of titanium under different conditions were calculated according to the conventional cross-sectional metallographic examination method. Those results suggested that the thickness of the materials is dependent on the electric parameters, time and the environmental factors.

OPEN IN VIEWER

Table 2

General properties of ESD on titanium alloy surface

ESD layer	Thickness/μm	Composition	Roughness/μm
Nb in air	8–10	Nb, Ti	2·68
YG8 in air	15	W2C, Co3W3C, CoTi	1·57
Si in air	40	TiN, Ti5Si3, TiSi2	1·45
Si in silicone	32	TiC, Ti5Si3, TiSi2	0·71

Nb was deposited to form the mutual soluble layer, while the modification of the cemented carbide electrode is the formation of carbides and metal compounds. The main component of deposition layer is porcelain for Si electrode in the two media. The roughness and the thickness of the two former are increased, which is attributed to the discharge parameters and discharge splash. However, the roughness of the latter two is decreased, due to the delay in semiconductor silicon discharge and the electrode pulverisation. The results also show that both N and Si are involved in the reaction under the air and silicone oil environment.

Figure 1b shows that the pollutants on the surface can be effectively removed with Al₂O₃ sandblasting treatment. It is beneficial to bond for the titanium–porcelain. The competition between skinning and coursing could improve the effective area of mechanical bonding (the value of the R_a for the modified Si layer in the air decreased to 1·05 μm).

From the section of the ESD layer, the layer and the substrate are bound tightly, without obvious boundaries. Figure 2 shows the typical morphology of the Si layer profile. It is indicated that the strengthening layer combined well with the substrate. Figure 3 is the elemental distributions along the depth of ESD layers in silicone oil. The results show that the composition of the modified layer changes gradient, which shows that the modified layer has both the composition of substrate and electrode after the metallurgical reactions. The ESD processing integrates the deposition layer and the substrate into a new system, achieving the combination of the atoms, which is the guarantee for the titanium–porcelain bonding.

OPEN IN VIEWER

2

Imagine (SEM) of cross-section of specimen prepared by Si electrode in silicone oil medium

OPEN IN VIEWER

3

Elemental distributions along depth of ESD layer prepared by Si electrode in silicone oil media

Bond properties for titanium–porcelain

Figure 4 shows the bonding properties of the modified layer for the different depositions of titanium–porcelain. The bonding strength for Si in the air is optimality condition, which has improved significantly for titanium–porcelain bond strength (increased by 42). While the bonding strength for YG8 and Nb is similar with the Si in the air, which is improved ∼30 and 25 respectively. This trend is in good agreement with previous observations about the substrate binding strength. ¹⁷ Based on the results, the appropriate ESD can enhance the titanium–porcelain to meet the clinical needs.

OPEN IN VIEWER

4

Bond strength comparison of all specimens

Images (SEM) of the titanium/porcelain interfaces are shown in Fig. 5. The images of the cross-section of the tested base titanium–porcelain specimens demonstrated that the interface of the titanium specimen had less porosity than that of the sample sintered conventionally.

OPEN IN VIEWER

5

a titanium sintering treatment; b titanium debonding; c YG8 modification layer sintering treatment; d YG8 modification layer debonding; e Nb modification layer sintering; f Nb modification layer debonding; g Si modification layer prepared in air debonding

In fact, a white layer of Ti oxide may form at porcelain sintering temperature on uncoated titanium by physical examination. If removing the oxide with tweezers, it demonstrated that the oxide was extremely non-adherent. Figure 5a showed that the titanium surface, without using a bonding agent, had a poor metal–porcelain bond. The titanium–porcelain systems using modification such as YG8 and Nb showed better bonding and interfacial characteristics (Fig. 5d and f). The titanium–porcelain system exhibited an appearance of metal/porcelain interface similar to that of the titanium surface system only after sintering treatment (Fig. 5c and e). Lower bond strength between titanium and different porcelains was reported previously by others. In the present study, the bond strengths of the titanium–porcelain specimens are increased.

The characteristics of the residual porcelain on the separation interface are observed due to the similar properties of the roughness. From the section of ESD layer in the. 5g, the fracture mode for titanium–porcelain is the mixed fracture of adhesion and cohesion. The images of the titanium surface after debonding demonstrated residual porcelain retained on the metal surface for all groups. This observation indicates a combination of cohesive and adhesive bond failures. However, more traces of porcelain were observed on specimens that were treated with the combination of airborne particle abrasion and bonding agent, which may indicate that the failure was primarily adhesive for the rest of the groups.

Imaged of cross-sections of tested specimens demonstrated that the fracture line was located more often between ceramic and metal for all the groups except for specimens that were treated by the airborne particle abrasion and bonding agent combination, where the fracture line, in some areas, occurred primarily within the porcelain material.

Bond strength factors for titanium–porcelain

Based on the above analyses, it is obvious that for protective coating is essential to prevent the formation of excessive titanium–porcelain in order to maintain a compact bond of Ti and porcelain. Because titanium possesses large negative Gibbs energy of oxidation, continued oxidation could occur resulting from reduction of the oxides in the porcelain via displacement reactions. Therefore, porcelain induced oxidation at elevated temperature may result in a substantial decrease in oxide adherence to titanium. Up to now, there is limited information regarding the bonding strength between dental porcelain and titanium. Many researchers are concerned about the effective methods to improve the titanium–porcelain bond.

Improved titanium–porcelain bond strength on sandblasted titanium surfaces is the key to the surface engineering. Since the titanium can be involved by the substrate at the high temperature, titanium–porcelain compatibility is important in the making of ceramic–metal restorations.

In the studies on high temperature processing of titanium substrate and the interface coordination, the most important is how to inhibit the oxidation in the process of high temperature porcelain. Therefore, many methods are used to resist the oxidation, such as preoxidation, ¹⁸ the oxygen coatings (Au, Cr, NbN, TiAlN and ZrSiN) and microarc oxidation. Among them, the ESD layer of silicone oil in the preparation is better than the sedimentary layers in the air dense and integrity, which have better high temperature oxidation resistance. High temperature oxidation resistance of Nb and carbide modified layer is also good and thus effectively protects the oxidation of the titanium alloy substrate to reduce the bond strength. Figure 6 shows the high temperature oxidation behaviour comparison of ESD layers by Si electrode (750°C). The results indicate that the basis for the titanium–porcelain strong binding is the excellent antioxidant ability. Among them, the ESD layer prepared by Si oil shows the best antioxidant ability with the compactness and integrity, which can protect the titanium alloy substrate to reduce the bond strength.

OPEN IN VIEWER

6

Comparison of oxidation behaviour of ESD layers at 750°C

The oxide layer formed after pretreatment process gave rise to the occurrence of chemical bonding at the titanium–porcelain interface during sequential firing, thereby preventing the initiation of failure from taking place at this interface.

The results demonstrated that there was a strong chemical correlation between oxide adherence and area fraction of retained porcelain on the titanium at high temperature. When an oxide layer was of insufficient thickness to prevent complete dissolution by the fusing porcelain, the porcelain came into direct contact with the alloy surface, leading to poor adherence. Könönen and Kivilahti elucidated that the Si component of porcelain could diffuse through the thin oxide layer into the bulk titanium, whereby forming the brittle ESD layer of Ti₅Si₃ containing some oxygen. This situation resulted in the absence of porcelain remaining on the metal substrate.

The ESD layer could react to improve the interface effect, which is the similar process of the sintering on the porcelain compound and dental porcelains. The metallurgic reaction can make a great change of the interatomic fusion to enhance the bond strength for titanium–porcelain.

In order to realise the transition fusion of heterogeneous materials, we also study the effect of titanium–porcelain bonding on metals, porcelains and metal compounds containing porcelain material substances. It proved that the failure of the titanium–porcelain predominantly occurred at the titanium/oxide interface. The ESD layer was more strongly bonded to the porcelain than titanium. The poor adhesion of the ESD with the substrate was due to the thermal stress arising from large lattice mismatch and the large difference in coefficient of thermal expansion between Ti and layer during cooling.

In addition, the mechanical riveting cooperation is also able to help improve titanium–porcelain bond strength. The roughness of the ESD for Nb modified layer increases the effective area of the connection. The rough surface of titanium–porcelain by ESD modification is easy to form a chemical connection.